Are missing values important for earnings forecast? a machine learning perspective

Ajim Uddin, Chia-Ching Chou

Research output: Contribution to journalArticlepeer-review

Abstract

Analysts' forecast is one of the most common and important measures for firms' earnings prediction. However, it is a challenge to fully utilize it due to the missing values. This study applies machine learning techniques to impute missing values in individual analysts' forecasts and then predicts future earnings. After imputing missing values, the forecast error is reduced by 41%, suggesting that missing values contain useful information for earnings forecast. Forecast accuracy improves most for firms with high analysts' dispersion. Finally, we propose a stochastic gradient descent based coupled matrix factorization (CMF) to augment the imputation quality of missing values with firm characteristics. CMF can further reduce the error of earnings forecast by 19%.<br>
Original languageEnglish
JournalQuantitative Finance
StateAccepted/In press - 1800

Fingerprint

Dive into the research topics of 'Are missing values important for earnings forecast? a machine learning perspective'. Together they form a unique fingerprint.

Cite this